All Projects
EdTech / Adaptive Learning
EduSphere
Interactive learning platform with live classrooms and a recommendation model that personalizes each student's learning path in real time.
Most online learning platforms move every student through the same fixed curriculum at the same pace, which means a student who's already mastered a concept sits through it again, and one who's struggling gets pushed forward before they're ready — the platform has no way to tell the difference.
EduSphere was designed so the curriculum path itself is a variable, not a fixed sequence — live classrooms and content delivery all report back into a single model of what each student actually knows, which then shapes what that student sees next.
A recommendation model tracks each student's performance across exercises and live-session participation, builds a running estimate of their mastery per concept, and reorders upcoming lessons and practice material in real time — moving a student past material they've already mastered and inserting extra reinforcement exactly where their estimated mastery is weakest, rather than waiting for a unit test to reveal the gap.
Students spend their time on the material that's actually challenging for them instead of repeating what they already know, and instructors running live classrooms get a per-student mastery view that shows who needs attention before a test surfaces the problem.
EduSphere was designed so the curriculum path itself is a variable, not a fixed sequence — live classrooms and content delivery all report back into a single model of what each student actually knows, which then shapes what that student sees next.
A recommendation model tracks each student's performance across exercises and live-session participation, builds a running estimate of their mastery per concept, and reorders upcoming lessons and practice material in real time — moving a student past material they've already mastered and inserting extra reinforcement exactly where their estimated mastery is weakest, rather than waiting for a unit test to reveal the gap.
Students spend their time on the material that's actually challenging for them instead of repeating what they already know, and instructors running live classrooms get a per-student mastery view that shows who needs attention before a test surfaces the problem.
Project Details
Category
EdTech / Adaptive Learning
Completed
Aug 2025
Tech Stack
Vue.js
WebRTC
TensorFlow